BioWorld International Correspondent

LONDON - Genetically engineered bacteria, which invade cancer cells and deliver a cocktail of harmful proteins from within, might comprise the cancer therapy of the future.

A study in the UK suggests that strategy is possible and has confirmed it can have a therapeutic effect in a mouse model of cancer.

Georges Vassaux, principal investigator at the Cancer Research UK Molecular Oncology Unit of the Hammersmith Hospital in London, told BioWorld International: "We have shown that we can deliver active proteins to mammalian cells using bacteria as vectors. For treatment of cancer, these proteins can be enzymes that could metabolize nontoxic prodrugs into toxic components that would kill all the cells invaded in this way."

In experiments, Vassaux and colleagues showed that treating a mouse model of cancer with genetically engineered bacteria able to enter cancer cells, with later administration of a prodrug, could slow tumor growth.

Their study is reported in the April 22, 2004, issue of Gene Therapy in a paper titled "Potential therapeutic applications of recombinant, invasive E. coli."

The work has its roots buried in some unusual experiments carried out by physicians at the turn of the 20th century, who tried inoculating infectious bacteria, such as the organism that causes syphilis, into tumors. It was reported that there were sometimes dramatic results in which the tumor either shrank or disappeared.

Vassaux said: "Because there were obviously no clinical trials at that time, it is difficult to know whether anything rational was happening. And this work disappeared when chemotherapeutic agents began to be used against cancer. We started our project in order to reassess these ideas of whether injecting unpleasant infectious agents into tumors could destroy them, but using today's molecular tools."

Normally, Escherichia coli bacteria cannot enter human cells by passing through the outer membranes. To get around that problem, the researchers engineered a gene from Yersinia pseudotuberculosis into the genome of a nontoxic laboratory strain of E. coli. That gene encodes a protein called invasin, which binds to b1-integrin, a protein found on the surface of mammalian cells. Binding of b1-integrin stimulates the cells to phagocytose the bacteria.

In addition to the gene for invasin, the researchers inserted into the bacterial genome a gene from Listeria monocytogenes, which encodes a protein called LLO. Once the bacteria have entered the lysosome of the mammalian cell, LLO forms pores in the lysosomal membrane, releasing the bacterial contents into the cytoplasm of the mammalian cell.

One of the products of E. coli's own genes, an enzyme called purine nucleoside phosphorylase, is known to convert the harmless prodrug 6-methylpurine-2'-deoxyriboside (MPDR) into the toxin 6-methyl purine. Experiments showed that mammalian cells that had been invaded by the genetically engineered bacteria were much more sensitive to MPDR than those cells that had not been invaded. More than 90 percent of cells invaded by bacteria died when exposed to MPDR, compared to less than 15 percent of non-invaded cells.

When the researchers injected the genetically engineered bacteria into tumors in a mouse model, and then gave MPDR intraperitoneally, there was a significant inhibition of tumor growth 18 days later, compared with tumors in animals that had not been treated.

The group also proved that if the bacteria did not express the invasin gene, they would be cleared from the tumor following injection. Vassaux said: "We showed that those that express invasin are taken into the cancer cells, and can still be detected four days later. In the clinical situation, this could give you a therapeutic window of at least two days to administer your prodrug."

Next, the group plans to examine whether the immune response triggered by injection of the engineered cells into a tumor on one side of an animal will have any effect on a tumor on the other side of the same animal. Vassaux said: "We could see that dendritic cells and macrophages were attracted to the tumors following the treatment. We hypothesize that the treatment stimulates a strong immune response to the cells in the tumor and we want to investigate whether this could be enough by itself to eliminate the cells to which we deliver the therapeutic protein."